Accepted Manuscript

Title: Controlled synthesis of graphene via electrochemical

route and its use as efficient metal-free catalyst for oxygen
reduction

Authors: Nathanael Komba, Qiliang Wei, Gaixia Zhang,

Federico Rosei, Shuhui Sun

PII: S0926-3373(18)31041-5
DOI: https://doi.org/10.1016/j.apcatb.2018.10.070
Reference: APCATB 17158

To appear in: Applied Catalysis B: Environmental

Received date: 19-8-2018

Revised date: 22-10-2018
Accepted date: 27-10-2018

Please cite this article as: Komba N, Wei Q, Zhang G, Rosei F, Sun S, Controlled
synthesis of graphene via electrochemical route and its use as efficient metal-
free catalyst for oxygen reduction, Applied Catalysis B: Environmental (2018),
https://doi.org/10.1016/j.apcatb.2018.10.070

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
Controlled synthesis of graphene via electrochemical route and its use as efficient metal-free

catalyst for oxygen reduction

Nathanael Komba, Qiliang Wei, Gaixia Zhang, Federico Rosei*, Shuhui Sun*

PT
RI
Institut National de la Recherche Scientifique-Énergie Matériaux et Télécommunications,

SC
Varennes, QC J3X 1S2, Canada.

U
*
N
Corresponding e-mails: rosei@emt.inrs.ca (F.R); shuhui@emt.inrs.ca (S.S)
A
M

Graphical abstract
D
TE
EP
CC
A

1
Highlights:

 A cheap, fast and green chemistry strategy has been developed to fabricate high-quality
graphene.
 The number of layers of the graphene could be well controlled.
 N-doping places an important role in enhancing the ORR performance of graphene

PT
catalysts.
 The relationship between structure-performance of graphene based catalyst has been

RI
systematically studied and optimized.
 The N-doped graphene exhibits comparable activity and much better stability for ORR than

SC
Pt/C.

U
N
Abstract: The oxygen reduction reaction (ORR) is a critical step in the operation of fuel cells. The
A
exploration of efficient, low cost and durable non-precious ORR catalysts, to replace the rare and
M

expensive Pt-based catalysts, is promising towards the large-scale commercialization of fuel cells.
D

Here, we report the fabrication of high-quality graphene (GP) with controllable layers, using a
TE

simple, versatile and cost-effective electrochemical exfoliation method. Our systematic studies

revealed that the graphite precursor, electrolyte temperature, and pyrolysis conditions largely affect
EP

the number of layers and quality of the graphene, which consequently has a significant effect on its

ORR performance. The optimized graphene sample was subsequently heat-treated in the NH3
CC

atmosphere to obtain the nitrogen-doped graphene (N-GP). As a metal-free electrocatalyst, N-GP

A

exhibit much enhanced ORR activity than GP, and comparable ORR activity to the state-of-the-art

Pt/C catalyst. Moreover, it also shows better stability than Pt/C catalyst in alkaline solution.

2
1. Introduction

The oxygen reduction reaction (ORR) is a fundamental process in energy conversion devices

such as fuel cells [1] and energy storage technologies such as metal-air batteries [2,3]. The ORR is

a cathodic process which plays a critical role in the overall performance of a fuel cell. In aqueous

PT
solution, the ORR is a slow and highly irreversible process [4,5]. Due to the inherent nature of the

sluggish kinetics of ORR, the demand for highly efficient electrocatalysts is indispensable to

RI
improve the performance of fuel cells.

SC
Currently, platinum (Pt) based catalysts are the only commercially available efficient catalysts

for fuel cells. However, the high cost and low abundance of Pt hinder the wide-spread

U
commercialization of hydrogen fuel cells [6,7]. In addition, in normal test conditions, Pt-based
N
catalysts suffer from susceptibility to methanol cross-over, carbon monoxide poisoning, carbon
A
corrosion and Pt particles agglomeration [8]. Considerable efforts have been directed towards 1)
M

decreasing Pt content in the catalysts by developing either Pt-metal alloy or novel support
D

materials and 2) replacing Pt-based catalysts by either non-precious metal [9] or metal free
TE

materials [10-12]. In particular, the exploration of highly efficient and inexpensive metal-free

electrocatalysts for the ORR is being widely investigated [13]. Among various alternatives,
EP

graphene is an emerging and promising candidate for a metal-free catalyst to replace Pt, especially
CC

in alkaline media.

Graphene, a two-dimensional sheet of carbon atoms covalently bonded into a hexagonal lattice
A

matrix, displays unique physicochemical properties, such as large surface area, high electronic

conductivity, and excellent chemical stability [14-16]. These extraordinary features may be useful

for many potential applications such as electronic devices, transparent electrodes, catalysis, energy

storage and conversion [17-19]. Moreover, it has been shown that the catalytic activity of graphene

3
can be increased significantly by doping it with heteroatoms such as N and S atoms [15,16,20,21].

Accordingly, when graphene is doped with N-atoms, it is found to have excellent electrocatalytic

activity towards ORR in alkaline media.

The specific method used to prepare graphene is also important for ORR electrocatalysis, as it is

PT
closely related to the quality, yield, and efficiency of the final product, as well as its production

cost. Therefore, methods that preserve the inherent properties of graphene, while avoiding the use

RI
of any hazardous chemical/condition, are highly desired [22,23]. However, in the most traditional

SC
Hummer’s method for the preparation of graphene, strong mineral acids and oxidizing agents are

used, causing several manufacturing challenges which are not easy to address [24,25]. On the other

U
hand, for ORR applications, catalysts with high surface area, excellent electronic conductivity, and
N
balanced, functional groups are highly desired. Although perfect reduced graphene oxide (r-GO)
A
inherently has high surface area and excellent conductivity, its hydrophobic nature causes
M

challenges such as electrode wettability, thereby decreasing the ORR performance [26]. On the
D

other hand, the highly hydrophilic graphene oxide shows poor conductivity, which is unfavorable
TE

for ORR. Therefore, a method that could lead to high-quality graphene with well-balanced

catalytic properties is highly desirable. In this context, electrochemical exfoliation is a promising

EP

method for the preparation of graphene for ORR electrocatalysis. The electrochemical approach

potentially has many advantages over traditional chemical methods, including its being simple,
CC

easily scalable, and cost-effective, and most importantly allowing the tuning of the level of

oxidation [25,26]. Graphene exfoliated using electrochemical methods are of high quality, with a
A

few defects.

In this work, taking advantage of the electrochemical exfoliation method, we use two types of

raw materials, i.e., graphite plate and graphite foil, to prepare high-quality graphene under

4
different reaction conditions. Systematic studies revealed that N-doping was successfully achieved

with pyridinic nitrogen dominance. The ORR activity of the catalysts varied significantly with the

exfoliating and post-exfoliation conditions. The optimal N-doped graphene exhibits comparable

ORR performance to commercial Pt/C catalysts. Our work suggests that the low cost and

PT
environmentally friendly electrochemical method is a potentially viable approach to prepare

graphene with excellent ORR activity.

RI
SC
2. Experimental

2.1 Materials.

U
Graphite plate (99%) and foils (0.5 mm, 99.8%) were purchased from Alfa Aesar, Fischer
N
Scientific. Ammonium sulfate ((NH4)2SO4, 99 %)), Polyvinylpyrrolidone (average molecular
A
weight 40,000) (99%), 2,2,6,6-Tetramethyl-1-piperidinyloxy, (TEMPO) (98%), platinum on
M

Vulcan (20%, XC-72R; E-TEK), potassium hydroxide (KOH) and reagent alcohol, were purchased
D

from Sigma-Aldrich. All chemicals were of analytical grade and used without further purification.
TE

2.2 Electrochemical exfoliation and nitrogen functionalization of graphene.

EP

The exfoliation of graphene (GP) was carried out using a DC power supply in a two-electrode
CC

cell as reported elsewhere with slight modification [22, 24, 25]. In a typical synthesis, a graphite

substrate served as a positive electrode, and Pt wire acted as a negative electrode in an

A

electrochemical cell filled with (NH4)2SO4 (0.1 M) aqueous solution. A constant biased potential of

±10 V was applied between the two electrodes. The exfoliation of graphene was achieved at room

temperature (~22 °C) and in some cases carried out at different temperatures (22, 40, 60, and 80

°C) in a heated water bath. The duration of the exfoliation process was found to depend on the

5
graphite source, the concentration of the electrolyte and the applied bias potential. To gain insights

on the quality of the exfoliated graphene, different doses of an organic reagent, TEMPO was added

to some recipes. The exfoliation scheme of graphite is shown in Fig. 1. The as-exfoliated product

was collected, washed thoroughly with deionized water and reagent alcohol by vacuum filtration,

PT
and then dispersed in water by sonication for 15 minutes. The graphitic particles which are not

exfoliated were separated from graphene by centrifuging the dispersion at 1000 rpm for 10

RI
minutes. The fine suspension in the upper part of the centrifuge tube was fractionated into

SC
components with different graphene layers by controlling the centrifugation speed from 2500 rpm

(to get multi-layers), 5000 rpm (to get thin layers) to 14000 rpm (to obtain very thin layers). The

U
graphene powders were obtained by freeze-drying. The obtained graphene was functionalized with

N
nitrogen atoms by heating the graphene powders at 950 °C in argon for 60 minutes and then in
A
ammonia (NH3) for 15 minutes to obtain nitrogen-doped graphene (N-GP).
M
D

2.3 Physical characterization.

TE

The structural morphologies and composition of the catalysts were characterized using

transmission electron microscopy (TEM, JEOL JEM-2100F operated at 200 kV), field emission
EP

scanning electron microscopy (FE-SEM) and X-ray photoelectron spectroscopy (XPS) using a VG
CC

(Escalab220i - XL) equipped with Al Kα line as an excitation source. The electronic structure was

investigated using a UV-vis absorption spectrophotometer (thermos scientific Nanodrop 2000c).

A

2.4 Electrochemical characterization.

The catalyst ink was prepared by dispersing 10 mg of the catalyst to the solution of ethanol

(350 l) and Nafion ionomer (95 l, 5 wt %), followed by two cycles of 15 minutes sonication and
6
agitation in a vortex shaker. This ink, then casted on the disc part of the rotating ring disc electrode

(RRDE, 5.61 mm diameter glassy carbon disk and Pt ring) or the 5.0 mm diameter glassy carbon

rotating disc electrode (RDE), with a catalyst loading of 0.8 mg cm-2. For comparison, 20 wt%

Pt/C catalyst (E-ETK) was prepared through the same procedure with a loading amount of 100

PT
μg/cm2 (20 μgPt/cm2). All electrochemical measurements were carried out in a three-electrode cell

at room temperature using the Autolab electrochemical system (Model, PGSTAT-302, Ecochemie,

RI
Brinkman Instruments). In all experiments, Pt wire and mercury/mercury oxide (Hg/HgO) was

SC
used as the counter electrode (CE) and a reference electrode (RE), respectively. The ORR activity

of the graphene catalysts was evaluated in O2 saturated 0.1 M KOH solution with scan rates of 50

U
mV s-1 and 10 mV s-1 for cyclic Voltammetry (CV) and linear sweep voltammograms (LSVs),

N
respectively, recorded in the range of 0 – 1.2 V. To record LSVs, the electrode was rotated at 1600
A
rpm. The potentials in this work are normalized to the reversible hydrogen electrode (RHE). The
M

peroxide yield (H2O2%) and the electron transfer number (n) were calculated using a ring, disk

currents (Ir, Id) and the estimated collection efficiency (N = 0.37) using equation (1) and (2). To
D

investigate the stability of the catalysts accelerated degradation testing (ADT) were performed in
TE

the oxygen-saturated 0.1 M KOH between 0.6 and 1.0 V vs. RHE at 100 mV s-1. The LSV tests
EP

were carried in the oxygen-saturated electrolyte at the beginning and end of each ADT to evaluate

the effect of the cycles on catalytic activity.

CC

200  I r ………… (1)

H 2O2% 
N Id  Ir
A

4 Id …………… (2)
n 
Id  Ir N

7
3. Results and discussion

It is widely accepted that a complete electrochemical exfoliation of graphite begins with the

oxidation of graphite edges and grain boundaries, followed by intercalation of water molecules and

PT
electrolyte ions when a bias potential is applied [22, 25, 27, 28]. The electrolysis of the intercalated

RI
species produces some gaseous products which cause the expansion of graphitic layers and

consequently separation into particles with loosely attached layers or individual sheets of graphene

SC
[27, 28]. However, the mechanism of the exfoliation steps could be slightly different depending on

U
the type of the graphite raw material. For example, as shown in Fig. S1, during the exfoliation of

N
graphite foils, we observed only the expansion of graphite foil using a static bias potential of -10 V
A
for five minutes. For both processes, expansion, and exfoliation were observed when the bias
M

potential was +10 V for similar duration. This observation can be accounted by the degree of

compaction of the two kind of graphite substrates and the mechanism of exfoliation. The graphite
D

plate is made of fine grain, with high strength and density graphite as well as high degree of
TE

compaction, it is very hard compared to the soft nature of the graphite foils. In the mechanism of

graphite exfoliation, there are three suggested key processes; (i) oxidation of the edge sites and
EP

grain boundaries, (ii) intercalation of ions within the graphitic layers, and (iii) expansion and
CC

separation [25, 28]. Generally, in the optimized conditions of exfoliation of graphite foils, we

applied a bias potential of +2.5 V for wetting, then ramped up to +10 V for oxidation of edges,
A

intercalation, and exfoliation. Afterwards, -10 V is applied for the reduction of oxidized groups

[25]. Due to the soft nature of the graphite foil, during the wetting process (by application of +2.5

V), some of the electrolytes ions and water molecules easily intercalate into the layers of graphite

foils. Since the graphite edges are not oxidized, when a bias potential is changed to -10 V, the

8
reduced gaseous products of the electrolysis only causes the expansion of the graphite foil, but the

force is not enough to cause separation of the graphene sheet. However, when a bias potential of

+10 V is applied after wetting, all three key processes happen in sequence, i.e., the exfoliation is

initiated by oxidation of graphite edges, followed by the intercalation, expansion (electrolysis of

PT
gaseous products) and separation to graphene. On the other hand, the alternation of -10/+10 V

voltages was necessary to achieve the exfoliation of graphite sheets using a graphite plate as a

RI
precursor. The difference in the mechanism of exfoliation is probably caused by the different

SC
compact degree and level of order of the graphite used.

U
The morphologies of as-synthesized graphene samples were investigated by TEM. Fig. 2 shows

N
typical TEM images of graphene exfoliated from a graphite foil (Fig. 2a-c) and graphite plate (Fig.
A
2d-f). The graphene sheets from both graphite sources appear to be uniform and very thin, about
M

two to a few layers, as observed by counting the folded edges of the HRTEM images (Fig. 2b, e).

The higher magnification images in Fig. 2 (b, e) indicate that the graphene sheets are very thin, and
D

the folded edges reveal that the graphene is only a few layers. Fig. 2 (c) and (f) are the
TE

corresponding selected area electron diffraction (SAED) patterns, indicating the single crystalline
EP

lattice structure with a characteristic hexagonal diffraction pattern of few layers graphene obtained

from both graphite sources [27,29].

CC

However, there are some slight differences when using different graphite precursors. For
A

example, very uniform, large piece and ultrathin (1-2 layers) graphene sheets are obtained when

the graphite foil is used, while smaller pieces with non-uniform thickness graphene sheets are

obtained when the graphite plate is used (as shown in Fig. 2a and 2d). The degree of crystallinity

is higher for the graphene that is exfoliated from graphite foil than the graphite plate. This could be
9
attributed to the longer exfoliation time when the graphite plate is used, compared to graphite foils.

The exfoliation process, using a graphite plate as precursor took about one hour or longer to obtain

a significant amount of graphene produce, while it only takes 3-5 minutes to complete the process

when using a graphite foil. The long exfoliation time of graphite plate imparts the as-exfoliated

PT
graphene to a longer oxidizing environment, resulting in the formation of relatively small pieces

with a large number of oxidized functional groups. The selected area electron diffraction (SAED)

RI
features with less well-defined diffraction spots (Fig. 2f), confirms the poor crystalline structure of

SC
the GPp [30]. The SEM images in Fig. S2 show that the surfaces of the as-exfoliated graphene are

thin and smooth with wrinkled features.

U
N
UV-vis absorption spectroscopy was performed to determine the extent of the oxidized carbon
A
atoms which occurred during the exfoliation process. Fig. 3 shows the UV-vis absorption spectra
M

of the graphene samples exfoliated in an electrolyte containing different doses of reducing agent,

TEMPO. The absorption peak of the as-exfoliated graphene in an electrolyte without TEMPO is
D

observed at around 257 nm, and gradually redshifted to about 270 nm as the amount of TEMPO
TE

increased from zero to 0.96 mmol. The observed shifts in the absorption peak can be attributed to
EP

the restoration of C-C sp2 hybridization within the graphene lattice structure. The absorption peak

maximum for the graphene oxide is reported to be around 231 nm while that of the reduced
CC

graphene oxide occurs at ca. 271 nm [22, 31]. The absorption in this region (270 nm) arises from

the excitation -plasmon resonance in the graphene structure [31-33]. Therefore, the occurrence of
A

the absorption peak at ca. 257-270 nm for the as-exfoliated graphene further confirms the

production of graphene with a controllable degree of oxidation by means of electrochemical

exfoliation methods.

10
 The chemical composition of the graphene samples was probed by X-ray photoelectron

spectroscopy (XPS). Fig. 4 displays the XPS spectra of as-exfoliated graphene (GPp) and nitrogen

doped graphene (N-GPp). The survey spectra are shown in Fig. 4a suggests that the as-exfoliated

graphene contains a strong and prominent C 1s peak and relatively small O 1s peak, indicating

PT
fewer oxygen-containing functional groups. The oxygen content in the GPp measured by XPS is 5

at% leading to a calculated ratio of C/O to be 19. The high ratio of C/O implies that the as-

RI
exfoliated graphene has a high degree of sp2 carbon [25, 32, 34]. On the other hand, the survey

SC
spectrum of the N-GPp indicates the presence of N 1s peak in addition to C 1s and O 1s peaks,

suggesting the successful doping of N atoms into the GPp during pyrolysis at 950 ºC in NH3 for 15

U
min. The surface concentration of C 1s, N 1s, and O 1s in the N-GPp as estimated is 96.7 at%, 2.0

N
at% and 1.5 at%, respectively. The decreased oxygen content for the N-GPp is attributed to the
A
decomposition of carboxyl functional groups during pyrolysis and replacement of oxygen by
M

nitrogen bonded to carbon. As shown in Fig. 4b, the presence of the C 1s peak with the main

component at 284.6 eV for C=C, minor peaks at 285.5 eV for C-O and 286.7 eV for C=O in the as-
D

exfoliated GPp is a signature of the low level of oxidized species in the graphene lattice structure
TE

[25,35]. The C 1s peak for N-GP (Fig. 4c) can be deconvoluted into four peaks, corresponding to
EP

C=C (284.6 eV), C=N (285.5 eV), C-N (286 eV), and C=O (287 eV), respectively. It is well

established that the charge distribution in the C atoms in the N-doped carbon matrix can be
CC

influenced by the neighboring N dopants, which serve as a potential active center for ORR [36]. In

addition, Fig. 4d shows the deconvolution of the N 1s peak into four components corresponding to
A

different N atom configurations. The components are assigned to pyridinic-N (398.3 eV, 71 at %),

pyrrolic-N (400.2 eV, 9 at %), graphitic-N (401 eV, 16 at %), and oxidized-N (402.7 eV, 4 at %).

The N-doped graphene contains the highest composition of pyridinic-N which contributes one 

11
electron to the  system of graphene. This phenomenon is believed to be able to enhance the ORR

activity of the N-doped graphene [19, 31, 37, 38].

The LSV experiments were performed to evaluate the ORR activity of the as-exfoliated

PT
graphene from graphite foil (GPf) and plate precursors (GPp) and NH3-treated graphene

(exfoliated from the graphite plate, N-GPp). The ORR activity of the catalysts was evaluated by

RI
the exhibition of the highest possible positive value of the onset potential, half-wave potential

SC
(E1/2) or current density at 0.80 V vs. RHE. Fig. 5 shows the typical ORR polarization curves, the

corresponding calculated electrons transfer number and percentage peroxide produced for the as-

U
exfoliated GPf, GPp and N-GPp catalysts. The GPp exhibited superior positive onset potential
N
(0.82 V vs. RHE), the half-wave potential (0.75 V) and current density at 0.80 V (0.34 mA cm-2) as
A
compared to 0.78 V, 0.67 V and 0.04 mA cm-2 respectively, of the GPf. The as-exfoliated GPp
M

displayed better ORR performance in 0.1 M KOH solution than the graphene prepared by other

methods, e.g., three-dimensional (3D) graphene [39,40], nanosheets by grinding high purity
D

graphite in ionic liquid [14,41], edge-functionalized graphene by ball milling [26], and reduced
TE

graphene oxide by Hummer’s method [42,43]. Detailed comparison is given in Table S1. The
EP

improved ORR performance of GPp is ascribed to the presence of numerous functionalized edges

available for the adsorption and dissociation of O2 in addition to the efficient electron transfer
CC

between the electrode and the high structural crystalline quality of its basal plane [36, 44-45]. The

abundance of the edges of the former possibly originates from the exfoliation process. The
A

duration of the exfoliation process for the graphite plate was long (one to a few hours), compared

to the graphite foil (3-5 min). The extended period of the exfoliation process subjects the edges of

graphene to further oxidation, consequently leading to the formation of more functional groups.

The functionalized edges of graphene render the hydrophilic nature of the modified electrode,
12
facilitating the diffusion of oxygen, thereby providing more active centers [26, 45-46]. The ORR

activity of the GPp is further enhanced by the pyrolysis of the sample in NH3 at 950 C (named as

N-GPp). The high-temperature pyrolysis in NH3 introduces N-atoms into the graphene lattice

matrix, which in turn causes the change of the charge distribution and electronic properties of the

PT
neighboring carbon atoms, hence increasing the probability of adsorption of oxygen molecules [19,

36]. The increased ORR activity of the N-GPp is evidenced by the more positive shifts of both

RI
half-wave (0.83 V vs. RHE) and onset (0.96 V) potentials, as well as the current density at 0.80 V

SC
(3.0 mA cm-2). The enhanced activity is also revealed by the increase of the electron transfer

number (above 3.5) and the relatively low production of hydrogen peroxide, H2O2, (below 15%) in

U
the potential range of 0.20–0.70 V, as seen in Fig. 5b. Therefore, it is apparent that graphite
N
precursors and pyrolysis of graphene samples in NH3 significantly influence the ORR activity of
A
the exfoliated catalysts.
M

Subsequently, we studied the temperature effect of the exfoliation electrolyte on the ORR
D

performance of the graphene samples. Fig. 6a shows digital photos of the exfoliation products in
TE

the electrolytes (upper) and on the graphite substrates (lower) after 5 minutes of exfoliation, at

different electrolyte temperatures. The exfoliation speed/efficiency is observed to decrease when

EP

the temperature increases. For example, at room temperature (22 C), the largest amount of
CC

graphene was produced (upper image in Fig. 6a). This can also be supported by the largest

consumption of the graphite substrate at 22 C while the least consumption occurs at 80 C
A

(lower image in Fig. 6a). After exfoliation at various temperatures, the as-synthesized GPp

samples were heat treated in Ar atmosphere at 1050 C for 1 hour and then in NH3 for 15 minutes,

to obtain N-GPp catalysts. The ORR performance of these N-GPp catalysts also decreases with the

increasing temperature of the electrolyte, as shown in Fig. 6 (b-d). Among all the N-GPp samples

13
obtained at different temperatures, the sample exfoliated at 22 ºC exhibited the best ORR activity

in terms of highest half-wave (0.82 V) and onset (0.94 V) potentials, electron transfer number

(above 3.8), and current density (3.0 mA/cm2) at 0.80 V, as well as the lowest peroxide yield

(below 11%). The performance parameters of all N-GPp catalysts are summarised in Table S2.

PT
The higher density of the defects on the graphene obtained at a lower temperature may be

responsible for its better ORR activity [47]. Our results are consistent with their observations.

RI
Furthermore, the graphene samples exfoliated at different temperatures display similar ORR

SC
kinetics as evidenced by the close Tafel slopes (Fig. S3) at the lower overpotential region. We

believe that the presence of a suitable number of defects in the graphene edges could improve the

U
hydrophilicity of the graphene, which, in turn, facilitates the adsorption of O-atoms in neighboring

N
C-atoms while maintaining excellent electronic conductivity.
A
In addition, we studied the influence of the number of graphene layers, the pyrolysis
M

temperature in NH3, the sonication time, and the concentration of exfoliation electrolyte on the
D

ORR activity of the graphene samples, and the results are summarised in Fig. 7. Before the ORR
TE

performance test, all the GPp samples underwent pyrolysis at 950 C in an argon atmosphere for

one hour and NH3 for 15 minutes, except for two samples (at 750 C and 1050 C, respectively) in
EP

Fig. 7b, forming N-GPp samples. Fig. 7a shows that the ORR activity of the catalysts follows the

following order: few layers N-GPp > thin layers N-GPp > multilayers N-GPp. The enhanced
CC

activity of few layers N-GPp originates from the abundance of edges sites (capable of promoting
A

the adsorption of O2 molecules) and efficient electron transfer between the electrode’s surface and

the edges [14]. To study the effect of the pyrolysis temperature in NH3, the thin layers GPp was

used as reference material and annealed in NH3 for 15 min at 750 C, 950 C and 1050 C,

respectively. As shown in Fig. 7b, the current density at 0.80 V significantly increased from 0.5

14
mA/cm-2 for the sample at 750 C to 1.9 mA/cm-2 for the sample at 1050 C. The increase in ORR

activity can be attributed to the fact that the higher N amount was doped into the graphene lattice

at higher NH3 pyrolysis temperature, providing more active sites [11]. Furthermore, the influence

of the sonication time was investigated as it is an integral step in the post-exfoliation treatment of

PT
the products. As shown in Fig. 7c, a significant enhancement of the ORR activity of the catalysts is

observed when the sonication was prolonged from 5 mins to 120 mins. The enhancement in the

RI
activity is caused by the introduction of more defects (active sites) in the graphene structure as the

SC
sonication time was extended. In addition, the concentration of exfoliation electrolyte also

displayed a significant influence in the ORR activity of the N-GPp samples. As shown in Fig. 7d,

U
the N-GPp derived from graphene exfoliated at 1.0 M (NH4)2SO4 exhibited superior ORR
N
activities than the one at 0.1 M (NH4)2SO4. We anticipate that at high concentration of the
A
electrolyte, the oxidation process of graphite edges is fast and the graphene produced is,
M

accordingly, have more edge defects which in turn acts as the active sites for the ORR.
D
TE

Finally, to investigate the stability of the N-GPp and 20% Pt/C catalysts towards ORR, ADT

was performed by continuous potential cycling between 0.6 and 1.0 V in O2-saturated 0.1 M KOH,
EP

for 5000 potential cycles with a scan rate of 200 mV s-1. As shown in Fig. 8, the N-GPp displayed
CC

exceptional durability. After 5000 consecutive CV cycles, the N-GPp catalysts showed almost

superimposed LSVs before and after the ADT cycling. In contrast, the half-wave potential of the
A

20%Pt/C electrode significantly reduced by 30 mV after 5000 potential cycles under identical

conditions. This observation suggests that N-GPp is more stable than the commercial Pt/C catalyst

for ORR in alkaline media, which can be attributed to the strong covalent bonds between the active

sites and the graphitic lattice, as well as the high corrosion resistance of graphene.

15
4. Conclusions and Perspectives

High-quality graphene with a controllable number of layers was successfully fabricated by a

facile and inexpensive electrochemical method, using a low-cost graphite foil and plates as

PT
precursors in an aqueous electrolyte. In general, both graphite precursors generated thin and highly

uniform graphene. We found that the ORR activity of graphene exfoliated from the graphite plate

RI
was higher than the graphene obtained from the graphite foil. Further, as-synthesized graphene

SC
samples were converted into N-doped graphene (N-GPp) after NH3 pyrolysis, which exhibits a

highly enhanced ORR performance compared to pristine graphene, due to N-doping. Moreover,

U
the N-GPp exhibited comparable ORR activity and better stability than the state-of-the-art

N
commercial Pt/C catalyst. In addition, we found that the ORR activity of exfoliated graphene is
A
influenced by the exfoliation and post-exfoliation parameters such as the temperature of the
M

exfoliation electrolyte, the number of layers of graphene, pyrolysis temperature and concentration

of the electrolyte. Therefore, it is possible to fine-tune the ORR activity of the graphene and N-
D

GPp by manipulating the experimental parameters. Thus, we believe the N-GPp which can be
TE

prepared by a cost-effective method using low-cost precursors is a potential candidate of ORR

EP

catalyst in an alkaline fuel cell.

CC

Author Contributions
A

The manuscript was written through contributions of all authors. All authors have given approval

to the final version of the manuscript.

Notes

The authors declare no competing financial interest.

16
Acknowledgments
S.S. and F.R. acknowledge individual Discovery Grants from the Natural Sciences and

Engineering Research Council of Canada (NSERC), as well as funding from the Fonds de la

Recherche du Québec sur la Nature et les Technologies (FRQNT), le Centre Québécois sur les

Materiaux Fonctionnels (CQMF), the Canada Foundation for Innovation (CFI). N.K. is grateful to

PT
the UNESCO Chair in Materials and Technologies for Energy Conversion, Saving and Storage

RI
(MATECSS) for a Ph.D. Excellence Scholarship.

SC
U
N
A
M
D
TE
EP
CC
A

17
References
[1] Song, C.; Zhang, J., Electrocatalytic Oxygen Reduction Reaction. 2008; pp 89-134.
[2] Wei, Q.; Fu, Y.; Zhang, G.; Sun, S. Curr. Opin. Electrochem., 2017, 4, pp. 45–59.
[3] Healy, J.; Hayden, C.; Xie, T.; Olson, K.; Waldo, R.; Brundage, M.; Gasteiger, H.; Abbott, J.,

PT
Fuel cells, 2005, 5, pp. 302-308.
[4] Ge, X.; Sumboja, A.; Wuu, D.; An, T.; Li, B.; Goh, F. W. T.; Hor, T. S. A.; Zong, Y.; Liu,

RI
Z., ACS Catal., 2015, 5, (8), pp. 4643-4667.
[5] Ly, Q.; Merinov, B. V.; Xiao, H.; Goddard, W. A.; Yu, T. H., J. Phys. Chem. C., 2017, 121,

SC
(44), pp. 24408-24417.
[6] Winter, M.; Brodd, R. J., Chem. Rev., 2004, 104, (10), pp. 4245-4270.
[7] Cui, X.; Yang, S.; Yan, X.; Leng, J.; Shuang, S.; Ajayan, P. M.; Zhang, Z., Adv. Funct.

U
Mater., 2016, 26, (31), pp. 5708-5717.
[8] N
Jin-Cheng, L.; Peng-Xiang, H.; Chang, L., Small, 2017, 13, (45), pp. 1702002.
A
[9] Zhang, J.; Xia, Z.; Dai, L., Sc. Adv., 2015, 1, (7).
[10] Wei, Q.; Zhang, G.; Yang, X.; Chenitz, R.; Banham, D.; Yang, L.; Ye, S.; Knights, S.; Sun,
M

S. ACS Appl. Mater. Interfaces, 2017, 9, (42), pp. 36944-36954.

[11] Gong, K.; Du, F.; Xia, Z.; Durstock, M.; Dai, L., Science, 2009, 323, (5915), pp. 760-764.
D

[12] Dai, L.; Xue, Y.; Qu, L.; Choi, H.-J.; Baek, J.-B., Chem. Rev., 2015, 115, (11), pp. 4823-
TE

4892.
[13] Yao, Z.; Yan, J.; Mietek, J.; Yonggang, J.; Zhang, Q. S., Small, 2012, 8, (23), pp. 3550-3566.
EP

[14] Benson, J.; Xu, Q.; Wang, P.; Shen, Y.; Sun, L.; Wang, T.; Li, M.; Papakonstantinou, P.,
ACS Appl. Mater. Interfaces, 2014, 6, (22), pp. 19726-19736.
CC

[15] Qu, L.; Liu, Y.; Baek, J.-B.; Dai, L., ACS Nano, 2010, 4, (3), pp. 1321-1326.
[16] Xuejun, Z.; Jinli, Q.; Lin, Y.; Jiujun, Z., Adv. Energy Mater., 2014, 4, (8), pp. 1301523.
[17] Tong, X.; Wei, Q.; Zhan, X.; Zhang, G.; Sun, S., Catalysts, 2017, 7, (1), p. 1.
A

[18] Ziyin, L.; Gordon, W.; Yan, L.; Meilin, L.; Ching-Ping, W. Adv. Energy Mater. 2012, 2, (7),
pp. 884-888.
[19] Wu, J.; Ma, L.; Yadav, R. M.; Yang, Y.; Zhang, X.; Vajtai, R.; Lou, J.; Ajayan, P. M. ACS
Appl. Mater. Interfaces, 2015, 7, (27), pp. 14763-14769.
[20] Wei, Q.; Tong, X.; Zhang, G.; Qiao, J.; Gong, Q.; Sun, S., Catalysts, 2015, 5, (3), p. 1574.

18
[21] Shuangyin, W.; Lipeng, Z.; Zhenhai, X.; Ajit, R.; Wook, C. D.; Jong-Beom, B.; Liming, D.,
Angew. Chem., Int. Ed., 2012, 51, (17), pp. 4209-4212.
[22] Parvez, K.; Li, R.; Puniredd, S. R.; Hernandez, Y.; Hinkel, F.; Wang, S.; Feng, X.; Müllen,
K., ACS Nano, 2013, 7, (4), pp. 3598-3606.
[23] Munuera, J. M.; Paredes, J. I.; Enterría, M.; Pagán, A.; Villar-Rodil, S.; Pereira, M. F. R.;
Martins, J. I.; Figueiredo, J. L.; Cenis, J. L.; Martínez-Alonso, A.; Tascón, J. M. D., ACS

PT
Appl. Mater. Interfaces, 2017, 9, (28), pp. 24085-24099.
[24] Na, L.; Fang, L.; Haoxi, W.; Yinghui, L.; Chao, Z.; Ji, C. Adv. Funct. Mater., 2008, 18, (10),

RI
pp. 1518-1525.
[25] Su, C.-Y.; Lu, A.-Y.; Xu, Y.; Chen, F.-R.; Khlobystov, A. N.; Li, L.-J., ACS Nano, 2011, 5,

SC
(3), pp. 2332-2339.
[26] Jeon, I.-Y.; Choi, H.-J.; Jung, S.-M.; Seo, J.-M.; Kim, M.-J.; Dai, L.; Baek, J.-B., J. Am.

U
Chem. Soc., 2013, 135, (4), pp. 1386-1393.

N
[27] Parvez, K.; Wu, Z.-S.; Li, R.; Liu, X.; Graf, R.; Feng, X.; Müllen, K., J. Am. Chem. Soc.,
2014, 136, (16), pp. 6083-6091.
A
[28] Yang, S.; Brüller, S.; Wu, Z.-S.; Liu, Z.; Parvez, K.; Dong, R.; Richard, F.; Samorì, P.; Feng,
M

X.; Müllen, K., J. Am. Chem. Soc., 2015, 137, (43), pp.13927-13932.
[29] Abdelkader, A. M.; Kinloch, I. A.; Dryfe, R. A. W. ACS Appl. Mater. Interfaces, 2014, 6,
D

(3), pp. 1632-1639.

TE

[30] Wang, G.; Yang, J.; Park, J.; Gou, X.; Wang, B.; Liu, H.; Yao, J., J. Phys. Chem. C 2008, 112
(22), 8192-8195.
[31] Li, D.; Müller, M. B.; Gilje, S.; Kaner, R. B.; Wallace, G. G., Nat. Nanotechnol., 2008, 3, p.
EP

101.
[32] Ganguly, A.; Sharma, S.; Papakonstantinou, P.; Hamilton, J., J. Phys. Chem. C, 2011, 115,
CC

(34), pp. 17009-17019.

[33] Saxena, S.; Tyson, T. A.; Shukla, S.; Negusse, E.; Chen, H.; Bai, J., Appl. Phys. Lett., 2011,
A

99, (1), p. 013104.

[34] Parvez, K.; Wu, Z.-S.; Li, R.; Liu, X.; Graf, R.; Feng, X.; Müllen, K., J. Am. Chem. Soc.,
2014, 136, pp. 6083-6091.
[35] Parvez, K.; Li, R.; Puniredd, S. R.; Hernandez, Y.; Hinkel, F.; Wang, S.; Feng, X., Müllen,
Klaus., ACS Nano, 2013, 7, (4 ), pp. 3598-3606.

19
[36] Flyagina1, I. S.; Hughes, K. J.; Mielczarek, D. C.; Ingham, D. B.; Pourkashanian, M., Fuel
Cells, 2016, 16, (5), pp. 568-576.
[37] Wu, J.; Ma, L.; Yadav, R. M.; Yang, Y.; Zhang, X.; Vajtai, R.; Lou, J.; Ajayan, P. M. ACS
Appl. Mater. Interfaces, 2015, 7, pp. 14763-14769.
[38] Xing, T.; Zheng, Y.; Li, L. H.; Cowie, B. C. C.; Gunzelmann, D., Qiao, Shi Zhang; Huang,
S.; Chen, Y., ACS Nano, 2014, 8, (7), pp. 6856–6862.

PT
[39] Wang, L.; Sofer, Z.; Ambrosi, A.; Šimek, P.; Pumera, M., Electrochem. Commun., 2014, 46,
pp. 148-151.

RI
[40] Lin, Z.; Waller, G. H.; Liu, Y.; Liu, M.; Wong, C.-p., Nano Energy, 2013, 2, (2), pp. 241-
248.

SC
[41] Deng, D.; Yu, L.; Pan, X.; Wang, S.; Chen, X.; Hu, P.; Sun, L.; Bao, X., Chem. Commun.,
2011, 47, (36), pp. 10016-10018.

U
[42] Bikkarolla, S. K.; Cumpson, P.; Joseph, P.; Papakonstantinou, P., Faraday Discuss., 2014,
173, (0), pp. 415-428. N
[43] Parvez, K.; Yang, S.; Hernandez, Y.; Winter, A.; Turchanin, A.; Feng, X.; Müllen, K., ACS
A
Nano, 2012, 6, (11), pp. 9541-9550.
M

[44] Yuan, W.; Zhou, Y.; Li, Y.; Li, C.; Peng, H.; Zhang, J.; Liu, Z.; Dai, L.; Shi, G., Sci. Rep.,
2013, 3, pp. 2248.
D

[45] Ikeda, T.; Hou, Z.; Chai, G.-L.; Terakura, K., J. Phys. Chem. C, 2014, 118, (31), pp.17616-
TE

17625.
[46] Xu, Z.; Fan, X.; Li, H.; Fu, H.; Lau, W. M.; Zhao, X., Phys. Chem. Chem. Phys., 2017, 19,
(31), pp. 21003-21011.
EP

[47] Hossain, S. T.; Wang, R., Electrochim. Acta, 2016, 216, pp. 253-260.
CC
A

20
Figure captions:

PT
RI
SC
U
N
Fig. 1. Schematic illustration of the setup for the electrochemical exfoliation production of
A
graphene, and the exfoliation mechanism in aqueous (NH4)2SO4 solution.
M
D
TE
EP
CC
A

21
 PT
RI
SC
U
N
A
Fig. 2. TEM images and selected area electron diffraction (SAED) of typical graphene samples
M

exfoliated from graphite foils and plates. (a-c) Graphite foil graphene (GPf). (d-e) Graphene of
D

graphite plate (GPp). The graphene from both graphite sources are few-layer, uniform, and have a
TE

well-defined crystalline structure.

EP
CC
A

22
 PT
RI
SC
U
N
A
Fig. 3. UV-vis absorption spectra are showing absorption peaks of the graphene samples prepared
M

in 0.1 M (NH4)2SO4 electrolyte containing different doses of organic reducing agent 2,2,6,6-
D

Tetramethylpiperidine-1-oxyl (TEMPO). The reducing agent was added to reduce part of the
TE

functional groups introduced during exfoliation.

EP
CC
A

23
 PT
RI
SC
U
N
A
M

Fig. 4. XPS spectra of graphene exfoliated from a graphite plate (GPp) (a) survey of as-exfoliated

graphene (GP) and N-doped graphene (N-GPp) and high-resolution C 1s spectrum of (b) as-
D

exfoliated GPp, and (c) N-GPp, and (d) high-resolution N 1s spectrum of the N-GPp.
TE
EP
CC
A

24
 PT
RI
SC
Fig. 5. ORR polarization curves of (a) as-exfoliated graphene from graphite foil (GPf, black line)

and graphite plate (GPp, red line), and N-doped graphene (N-GPp, blue line), (b) The calculated

U
electron transfer number, and peroxide production in the three catalysts. The measurements were

N
carried out in N2, or O2 saturated 0.1 M KOH, scan rate: 10 mV/s at 1600 rpm.
A
M
D
TE
EP
CC
A

25
 PT
RI
SC
U
N
A
M

Fig. 6. (a) Digital images of as-exfoliated graphene products (in the beakers) and their
D

corresponding graphite substrates after 5 minutes of exfoliation at 22, 40, 60 and 80 ºC,
TE

respectively. (b) ORR polarization curves ranging from 0 – 1.0 V of the graphene (N-GPp)

exfoliated at four different electrolyte temperatures, followed by pyrolysis in Ar at 1050 C for 1

EP

hour and NH3 for 15 minutes. The corresponding calculated (c) electron transfer number and (d)
CC

peroxide produced for the catalysts. Electrolytes: 0.1 M KOH, scan rate: 10 mV/s and rotation:

1600 rpm.
A

26
 PT
RI
SC
U
N
A
M
D

Fig. 7. The LSVs curve of N-GPp catalysts. (a) with different layers, (b) after different
TE

temperatures of pyrolysis in NH3 (c) sonication for different duration. (d) Exfoliated in 0.1 M and

1.0 M (NH4)2SO4 electrolytes. All LSV measurements were carried out in 0.1 M KOH media, scan
EP

rate: 10 mV/s, rotation speed: 1600 rpm.

CC
A

27
 PT
RI
SC
Fig. 8. Long-term performance of the catalysts (before and after ADT test, 5000 cycles). LSVs (a)

few layer N-GPp and (b) 20% Pt/C recorded at 10 mV/s and1600 rpm in O2-saturated 0.1 M KOH

U
at room temperature.
N
A
M
D
TE
EP
CC
A

28